and Helminthostachys zeylanica. 43 
has been given off is at first solid (Fig. 67), though centrally 
situated parenchyma usually reappears at a slightly higher 
level. 
It depends on the individual plant how soon the stele begins 
to have a larger pith and approximate to the structure 
described and figured by Farmer and Freeman. The simple 
type just described may persist for a number of internodes. 
The material did not afford stages old enough to allow of the 
transition to the condition of things in the old rhizome being 
followed. The most advanced stage found was closely similar 
to the simplest stele described by Farmer and Freeman 1 . At 
this stage the stele as a whole was larger than that in Fig. 65 ; 
the pith consisted of a considerable number of cells as seen in 
cross section, though tracheides were distributed through it. 
No internal endodermis was present and the departing leaf- 
trace did not even at this stage leave a leaf-gap, the ring of 
xylem closing up immediately. 
In the solid steles and in those with a small pith the proto- 
xylem is at, or nearest to, the centre and the development of 
the xylem is centrifugal. The elements of the protoxylem are 
spiral while the later formed tracheides are pitted. This 
endarch position of the protoxylem persists for several inter- 
nodes in most cases (Fig. 70). Whether in the cases, in which 
tracheides were distributed through the pith, the protoxylem 
was mesarch or endarch could not be determined. 
The origin of the first leaf-trace, with which succeeding 
traces originating from the more or less solid stele agree, is 
represented in Figs. 67 and 68. At the level at which the 
Sections figured are taken the endodermis is still complete, and 
encloses the stele together with the leaf- trace. At a slightly 
higher level it ceases to be continuous and immediately closes 
around the stele ; it remains visible for some distance on the 
outside of the trace, and near the base of the latter forms 
a separate and complete sheath. The common type of leaf- 
trace is that shown in Fig. 67, which is throughout endarch. 
In one or two cases even the first trace was at its origin and 
1 Farmer and Freeman, loc. cit., Fig. 23. 
